Vibe Coding Purity Test

Check every item you've done.

  1. Posted a hot take thread about "SWE is dead" and meant it.
  2. Argued that "agents are just scripts" in replies.
  3. Said "prompts are the new programming language" unironically.
  4. Predicted the next model release month in public.
  5. Tweeted "evals are fake, vibes are real".
  6. Posted "context window is all you need".
  7. Dunked on a benchmark screenshot for being cherry-picked.
  8. Picked a side in open-source vs closed-model wars.
  9. Quoted-tweeted an AI doom thread with "skill issue".
  10. Posted a graph about exponentials to win an argument.
  11. Called a model launch a "nothing-burger" within an hour.
  12. Said "alignment is just product safety" to stir discourse.
  13. Started a "agents will replace PMs" debate.
  14. Tweeted "MCP is the new plugin ecosystem".
  15. Wrote "hot take: fine-tuning is dead".
  16. Complained about "AI slop" flooding the timeline.
  17. Ran a "doomer vs accelerationist" poll.
  18. Shared a "we're so back / it's so over" meme for a release.
  19. Claimed "RLHF ruined creativity" in a thread.
  20. Posted a prediction-market screenshot about AGI timelines.
  21. Made a "prompt injection will kill us all" take.
  22. Replied "benchmarks don't matter, users do".
  23. Declared a model "AGI-ish" and got ratioed.
  24. Live-tweeted a keynote to score points.
  25. Announced you're leaving AI Twitter, then returned next week.
  26. Tried ChatGPT for the first time
  27. Asked an LLM to explain a bug in your code
  28. Used AI to rewrite an email or message
  29. Copied a prompt from someone else and ran it
  30. Turned on AI autocomplete in your editor
  31. Installed Cursor just to test AI edits
  32. Used GitHub Copilot to finish a small script
  33. Asked Claude to summarize a long article or PDF
  34. Generated a README with AI
  35. Used AI to brainstorm product or feature names
  36. Used AI to turn notes into a checklist
  37. Let an AI refactor a function for readability
  38. Ran a "fix lint errors" prompt
  39. Generated unit tests with AI
  40. Used AI to draft a cold outreach message
  41. Tried v0 or similar to mock a simple UI
  42. Tried Replit Agent on a toy project
  43. Asked AI to compare two libraries
  44. Used AI to debug a Dockerfile
  45. Let AI write a SQL query for you
  46. Asked AI to explain MCPs in plain English
  47. Followed an AI demo video step-by-step
  48. Used AI to translate text into another language
  49. Asked AI for a study plan to learn a tech stack
  50. Generated a logo or image with AI
  51. Paid for Cursor Pro to lift composer limits
  52. Upgraded to Claude Pro the day you hit the free cap
  53. Bought ChatGPT Plus just for GPT-4 access
  54. Subscribed to Copilot just for inline completions
  55. Used Cursor Composer to rewrite an entire feature
  56. Paid for v0 to ship a landing page overnight
  57. Used Bolt to scaffold a full-stack app in an hour
  58. Let Replit Agent set up a database and deploy it
  59. Tried Windsurf because someone said it beats Cursor
  60. Applied to the Devin waitlist and checked email daily
  61. Ran Clawdbot or Moltbot on a repo-wide refactor
  62. Bought a second AI tool subscription in the same week
  63. Cancelled a model subscription after a price hike
  64. Blew through $100 in API credits in a weekend
  65. Set an API budget and still blew past it
  66. Had an AI API bill bigger than your phone bill
  67. Paid for a vector database plan for embeddings
  68. Rented a GPU on RunPod or Vast for local models
  69. Bought new hardware just to run a local model
  70. Paid for Vercel Pro to host an AI demo
  71. Bought a domain just for an AI side project
  72. Paid for a prompt marketplace template pack
  73. Spent $500 plus in a month on AI tooling
  74. Stacked three AI subscriptions at once
  75. Paid for an agent framework course to speed up work
  76. Explained to a partner what "agentic" means, twice.
  77. Made a shared prompt doc for your team and gave it a name.
  78. Turned a Jira ticket into a prompt and called it a spec.
  79. Asked an LLM to write your standup update.
  80. Started your day by asking a model for a plan.
  81. Kept a context.md file just for the model.
  82. Gave a model a "project tour" before asking for changes.
  83. Used a prompt to draft meeting notes from a call.
  84. Sent a friend a prompt template instead of explaining the idea.
  85. Explained "context window" at a family dinner.
  86. Told your team you were "pairing with Claude" today.
  87. Created a "prompt cookbook" and actually used it.
  88. Let an agent open a PR and wrote the description for you.
  89. Ran a "plan -> execute -> reflect" loop more than once in a day.
  90. Set up a Slack channel just for prompts.
  91. Shared a screen to show someone a Cursor flow.
  92. Asked a coworker to review a prompt like code.
  93. Apologized for a late reply because your agent was running.
  94. Added a "prompting guide" to your team wiki.
  95. Built a tiny script to feed repo context into a model.
  96. Wrote a "prompt contract" for how your team asks for changes.
  97. Explained why your "pair" does not take coffee breaks.
  98. Saved a model's system prompt as a permanent file.
  99. Used an LLM to triage bugs before the team did.
  100. Changed how you write commit messages because a model does the first pass.